The long term goal of this research has been to define the role of oval cells in hepatocarcinogenesis. Previous investigations have focused on the analysis of cell lineages operative in the genesis of hepatocellular carcinomas using monoclonal antibodies (MAb) recognizing oval cell and hepatocyte associated antigens. Comparative phenotypic analysis of endogenous and transplanted neoplastic cell populations have provided support for the notion that hepatocellular carcinomas can arise from oval cell as well as hepatocyte progenitors. In this renewal application, we will extend and refine three aspects of our previous investigations in an effort to gain further insight into the importance of abnormal differentiation in the process of hepatocarcinogenesis. In the first specific aim, bacterial or eukaryotic expression systems will be used to isolate cDNA clones for OC.2, BD.1, H.2 and Hbd1, 4 MAb-defined antigens that have been shown to be valuable markers for tracking cellular lineage relationships during carcinogenesis and liver development. cDNA libraries prepared in a phagemid insertion vector (LambdaZap) or in the pCDM8 expression plasmid and expressed in E. coli or COS cells, respectively, will be screened with monoclonal antibodies specific for each antigen. Alternatively, LambdaZap libraries will be screened using degenerate oligonucleotides synthesized on the basis of microsequence obtained from purified peptides. Sequencing of cDNA clones will provide information regarding structure and function, homologies to known proteins and a means to design probes that can be used to examine expression at the RNA level and identify homologues in human cells. In aim 2, efforts will be directed towards identifying surface components involved in signaling events that occur during differentiation of oval cells or tumor cells previously demonstrated to undergo extensive differentiation in vivo. To identify such antigens, monoclonal antibodies will be produced using a novel immunization protocol involving serial immunological reconstitution of tumor bearing nude mice with immunocompetent spleen cells. This approach allows for a measured delay in the initiation of the immune response, thus giving transplanted cells a chance to become established and progress to a desired stage of differentiation. MAbs produced in this manner will provide a means to classify and/or subfractionate oval cells and neoplastic cell populations according to their expression of surface antigens closely linked to differentiation. In aim 3, pathways of differentiation and progression during carcinogenesis will be mapped by analyzing the fate of endogenous or transplanted cells which carry unique retroviral integration sites. Integration sites in cells infected in vitro or in vivo with a replication incompetent retroviral vector encoding the neomycin resistance gene (neo) will be identified by using inverse PCR to amplify genomic sequences flanking the integration site. Identification of unique integration sites held in common between oval cells and hepatocytes, focal lesions, nodules or primary tumors will provide unequivocal evidence for precursor-product relationships.